Cloning the Self for Mental Well-Being: A Framework for Designing Safe and Therapeutic Self-Clone Chatbots

Paper Title

Cloning the Self for Mental Well-Being: A Framework for Designing Safe and Therapeutic Self-Clone Chatbots

Publication Info

• Topic area: Design and evaluation of AI-driven self-clone chatbots for mental well-being.
• Keywords: self-clone chatbots, mental health, AI ethics, therapeutic design, user-clone relationship, personalization, safety, cognitive-behavioral therapy, compassion-focused therapy, digital mental health.

Background and Problem

• Problem / challenge: Current mental health chatbots lack deep personalization and fail to fully leverage users' internal narratives. Self-clone chatbots, while promising, pose risks such as reinforcing negative self-perceptions, identity confusion, and ethical concerns.
• Significance: Self-clone chatbots could transform mental health support by externalizing inner dialogues, fostering self-awareness, and offering personalized, low-barrier interventions.
• Motivation and related work: Prior research has explored AI clones, LLM-based chatbots, and digital mental health tools, but lacks a comprehensive framework for designing safe, therapeutic self-clones. Existing systems often fail to balance personalization with safety, leaving gaps in user trust and long-term engagement.

Solution

• Proposed approach: A design framework for self-clone chatbots aimed at supporting mental well-being, focusing on therapeutic grounding, design dimensions, and safety/ethical considerations.
• Novelty:
  1. A structured framework integrating psychological theories (e.g., parts-based therapy, CFT, CBT) with design principles for self-clone chatbots.
  2. Empirical insights from interviews with 16 mental health experts and 6 prospective users.
  3. Identification of risks and ethical dilemmas unique to self-clones, with actionable safety guardrails.
• Procedure and key techniques:
  • Conducted semi-structured interviews with experts and non-experts, using illustrative scenarios to explore perceptions of self-clone chatbots.
  • Applied thematic analysis to synthesize findings into a framework with three components:
    1. Therapeutic grounding and user considerations.
    2. Design dimensions (e.g., clone persona, replication fidelity, user-clone relationship).
    3. Safety and ethical considerations.

Results

• Concrete findings:
  • Self-clones can externalize inner dialogue, support perspective-taking, and enhance self-awareness.
  • Risks include reinforcing negative self-perceptions, over-identification with clones, and privacy concerns.
  • Experts highlighted the importance of grounding in therapeutic models like IFS, CFT, and CBT.
• Advantage over baselines:
  • Offers deeper personalization and therapeutic alignment compared to generic mental health chatbots.
  • Provides a structured approach to mitigate risks and enhance user trust.
• Experiments / evaluation:
  • Interviews with 16 mental health experts (mean clinical experience: 9.12 years) and 6 non-expert users.
  • Scenarios explored therapeutic affordances, risks, and design trade-offs.
  • Thematic saturation achieved through iterative coding and comparison with existing literature.
• Limitations and future work:
  • Lack of hands-on experience with functional self-clone prototypes.
  • Focused on non-clinical contexts; clinical applications require further study.
  • Future work should explore richer modalities (e.g., voice, visual likeness) and long-term efficacy.

Summary

This paper introduces a design framework for self-clone chatbots aimed at supporting mental well-being, grounded in therapeutic models like IFS, CFT, and CBT. By externalizing inner dialogue and offering personalized support, self-clones show promise for enhancing self-awareness and introspection. However, risks such as reinforcing negative self-perceptions and ethical dilemmas around identity and privacy necessitate careful design. The framework provides actionable guidance for developers, clinicians, and policymakers, emphasizing safety, user agency, and alignment with therapeutic goals. Future work should focus on practical trials, clinical integration, and expanding to richer modalities.